Search

In recent years, with the development of society and the rapid development of the animation industry, people are paying more and more attention to and requirements for animation production. As an indispensable part of animation production, picture composition plays a major role in animation production. It can give full play to the application of color matching and light and shadow design and enhance the depth and space of the animation screen. Tone space conversion refers to the conversion or representation of color data in one color space into corresponding data in another color space. Its purpose is to distinguish and process color components such as hue and saturation in an image. This article first introduces the domestic and foreign research status of digital image preprocessing and analyzes the basic principles of several color space conversions in detail. Then, several color space conversion algorithms are studied, and the performance of the algorithms is compared and analyzed. The paper focuses on the hardware implementation and optimization of the algorithm for converting RGB color space into HSI color space to meet the real-time requirements. This article focuses on the mutual conversion between the RGB tone space and the HSI tone space and describes in detail how each color component in the HSI tone space is converted from the three RGB color components from a geometric perspective, and then the conversion is derived, and several general conversion methods of RGB to HSI tone space are introduced; two conversion methods of geometric derivation method and standard modulus algorithm are implemented in the software, and the comparison verification is carried out, and the comparison is made from the perspective of hardware implementation. The pros and cons of the two methods are discussed. Finally, the paper summarizes the shortcomings in the design and proposes further research directions in the future. [ABSTRACT FROM AUTHOR]

Focusing on DOTA, the multidirectional object dataset in aerial view of vehicles, CMDTD has been proposed. The reason why it is difficult for applying the general object detection algorithm in multidirectional object detection has been analyzed in this paper. Based on this, the detection principle of CMDTD including its backbone network and multidirectional multi-information detection end module has been studied. In addition, in view of the complexity of the scene faced by aerial view of vehicles, a unique data expansion method is proposed. At last, three datasets have been experimented using the CMDTD algorithm, proving that the cascaded multidirectional object detection algorithm with high effectiveness is superior to other methods. [ABSTRACT FROM AUTHOR]

Aiming at the problem of low accuracy of edge detection of the film and television lens, a new SIFT feature-based camera detection algorithm was proposed. Firstly, multiple frames of images are read in time sequence and converted into grayscale images. The frame image is further divided into blocks, and the average gradient of each block is calculated to construct the film dynamic texture. The correlation of the dynamic texture of adjacent frames and the matching degree of SIFT features of two frames were compared, and the predetection results were obtained according to the matching results. Next, compared with the next frame of the dynamic texture and SIFT feature whose step size is lower than the human eye refresh frequency, the final result is obtained. Through experiments on multiple groups of different types of film and television data, high recall rate and accuracy rate can be obtained. The algorithm in this paper can detect the gradual change lens with the complex structure and obtain high detection accuracy and recall rate. A lens boundary detection algorithm based on fuzzy clustering is realized. The algorithm can detect sudden changes/gradual changes of the lens at the same time without setting a threshold. It can effectively reduce the factors that affect lens detection, such as flash, movies, TV, and advertisements, and can reduce the influence of camera movement on the boundaries of movies and TVs. However, due to the complexity of film and television, there are still some missing and false detections in this algorithm, which need further study. [ABSTRACT FROM AUTHOR]

Parallel processing as a method to improve computer performance has become a development trend. Based on rough set theory and divide-and-conquer idea of knowledge reduction, this paper proposes a classification method that supports parallel attribute reduction processing, the method makes the relative positive domain which needs to be calculated repeatedly independent, and the independent relative positive domain calculation could be processed in parallel; thus, attribute reduction could be handled in parallel based on this classification method. Finally, the proposed algorithm and the traditional algorithm are analyzed and compared by experiments, and the results show that the proposed method in this paper has more advantages in time efficiency, which proves that the method could improve the processing efficiency of attribute reduction and makes it more suitable for massive data sets. [ABSTRACT FROM AUTHOR]

This paper presents an in-depth study and analysis of Chinese painting image classification by a multitask joint sparse representation algorithm for texture feature extraction of Chinese painting images and proposes a method to extract texture features directly for the original images. It simplifies the process of image grayscale conversion and preserves the information contained in the original Chinese painting images to the greatest extent. The algorithm uses the ideas of multicolor domain analysis and multiscale analysis, combined with the traditional grayscale coeval matrix to extract texture features. Experiments show that the multiscale grayscale cooccurrence matrix algorithm outperforms the traditional grayscale cooccurrence matrix algorithm and the color grayscale cooccurrence matrix algorithm. The discriminative ability of multiple features for target recognition is integrated by multitask learning, thus improving the robustness and generalization ability of the algorithm; meanwhile, the recognition accuracy is improved by using a two-level multitask learning mode to exclude the interference of a large number of irrelevant dictionary atoms. The experimental results show that the algorithm has higher recognition accuracy and better robustness than the existing sparse representation SAR target recognition algorithm. Configuration recognition experiments are conducted on different configurations of target data, and the experimental results show that the algorithm achieves better configuration recognition accuracy than existing algorithms. [ABSTRACT FROM AUTHOR]

Aiming at the problems of low contrast and low definition of fog degraded image, this paper proposes an image defogging algorithm based on sparse representation. Firstly, the algorithm transforms image from RGB space to HSI space and uses two-level wavelet transform extract features of image brightness components. Then, it uses the K-SVD algorithm training dictionary and learns the sparse features of the fog-free image to reconstructed I-components of the fog image. Using the nonlinear stretching approach for saturation component improves the brightness of the image. Finally, convert from HSI space to RGB color space to get the defog image. Experimental results show that the algorithm can effectively improve the contrast and visual effect of the image. Compared with several common defog algorithms, the percentage of image saturation pixels is better than the comparison algorithm. [ABSTRACT FROM AUTHOR]

As the Internet of Things (IoT) smart mobile devices explode in complex opportunistic social networks, the amount of data in complex networks is increasing. Large amounts of data cause high latency, high energy consumption, and low-reliability issues when dealing with computationally intensive and latency-sensitive emerging mobile applications. Therefore, we propose a task-sharing strategy that comprehensively considers delay, energy consumption, and terminal reputation value (DERV) for this context. The model consists of a task-sharing decision model that integrates latency and energy consumption, and a reputation value-based model for the allocation of the computational resource game. The two submodels apply an improved particle swarm algorithm and a Lagrange multiplier, respectively. Mobile nodes in the complex social network are given the opportunity to make decisions so that they can choose to share computationally intensive, latency-sensitive computing tasks to base stations with greater computing power in the same network. At the same time, to prevent malicious competition from end nodes, the base station decides the allocation of computing resources based on a database of reputation values provided by a trusted authority. The simulation results show that the proposed strategy can meet the service requirements of low delay, low power consumption, and high reliability for emerging intelligent applications. It effectively realizes the overall optimized allocation of computation sharing resources and promotes the stable transmission of massive data in complex networks. [ABSTRACT FROM AUTHOR]

In this paper, we adapt the fractional derivative approach to formulate the flow-conservation transportation networks, which consider the propagation dynamics and the users' behaviors in terms of route choices. We then investigate the controllability of the fractional-order transportation networks by employing the Popov-Belevitch-Hautus rank condition and the QR decomposition algorithm. Furthermore, we provide the exact solutions for the full controllability pricing controller location problem, which includes where to locate the controllers and how many controllers are required at the location positions. Finally, we illustrate two numerical examples to validate the theoretical analysis. [ABSTRACT FROM AUTHOR]

Distinguishing target object under occlusions has become the forefront of research to cope with grasping study in general. In this paper, a novel framework which is able to be utilized for a parallel robotic gripper is proposed. There are two key steps for the proposed method in the process of grasping occluded object: generating template information and grasp detection using the matching algorithm. A neural network, trained by the RGB-D data from the Cornell Grasp Dataset, predicts multiple grasp rectangles on template images. A proposed matching algorithm is utilized to eliminate the influence caused by occluded parts on scene images and generates multiple grasp rectangles for objects under occlusions using the grasp information of matched template images. In order to improve the quality of matching result, the proposed matching algorithm improves the SIFT algorithm and combines it with the improved RANSAC algorithm. In this way, this paper obtains suitable grasp rectangles on scene images and offers a new thought about grasping detection under occlusions. The validation results show the effectiveness and efficiency of this approach. [ABSTRACT FROM AUTHOR]

In this paper, a new method of reactive power compensation is proposed for reducing power loss of distribution power networks. The new method is the combination of local compensation at each load and distribution line compensation. In the method, local capacitors at each load are determined to increase power factor of load to an expected value first and then a number of capacitors are placed in distribution lines with two factors, location and capacity by using the three applied methods. Particle swarm optimization (PSO), parasitism predation algorithm (PPA), and tunicate swarm algorithm (TSA) are applied for the proposed method on four distribution systems with 15, 33, 69, and 85 buses. The comparisons of results with previous methods indicate that the three applied methods can reach higher loss reduction for all study cases. TSA can reach loss reduction more effectively than others by 19.4%, 0.5%, 7.9%, and 10.76% for the four distribution systems, respectively. The most important emphasis is that PSO, which was considered to be of low effectiveness in previous studies, can reach much better loss than approximately all previous methods thank to the proposed compensation method. PSO also reached better loss reduction than others by 18.97%, 0.4%, 7.73%, and 10.21%, respectively. Consequently, it is recommended that the proposed compensation method is useful for distribution systems in minimizing total power loss and TSA can be selected a new and powerful method for the problem. [ABSTRACT FROM AUTHOR]

circRNA is a novel class of noncoding RNA with closed-loop structure. Increasing biological experiments have shown that circRNAs play an important role in many diseases by acting as a miRNA sponge to indirectly regulate the expression of miRNA target genes. Therefore, predicting associations between circRNAs and miRNAs can promote the understanding of pathogenesis of disease. In this paper, we propose a new computational method, NECMA, based on network embedding to predict potential associations between circRNAs and miRNAs. In our method, the Gaussian interaction profile (GIP) kernel similarities of circRNA and miRNA are calculated based on the known circRNA-miRNA associations, respectively. Then, the circRNA-miRNA association network, circRNA GIP kernel similarity network, and miRNA GIP kernel similarity network are utilized to construct the heterogeneous network. Furthermore, the network embedding algorithm is used to extract potential features of circRNA and miRNA from the heterogeneous network, respectively. Finally, the associations between circRNAs and miRNAs are predicted by using neighborhood regularization logic matrix decomposition and inner product. The performance of NECMA is evaluated by using ten-fold cross-validation. The results show that this method has better prediction accuracy than other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

The probability ranking conclusion is an extension of the absolute form evaluation conclusion. Firstly, the random simulation evaluation model is introduced; then, the general idea of converting the traditional evaluation method to the random simulation evaluation model is analyzed; on this basis, based on the rule of "further ensuring the stability of the ranking chain on the basis of increasing the possibility of the ranking chain," two methods of solving the probability ranking conclusion are given. Based on the rule of "further guaranteeing the stability of the ranking chain on the basis of improving the likelihood of the ranking chain," two methods are given to solve the likelihood conclusion. This paper argues that this absolute form of conclusion hinders the approximation of the theory to the essence of the actual problem and is an important reason for the problem of "non-consistency of multi-evaluation conclusions." To address this problem, a stochastic simulation-based comprehensive evaluation solution algorithm based on the idea of "Monte Carlo simulation" is proposed, and the corresponding ranking method is investigated, which is characterized by generating evaluation conclusions with probability (reliability) information, and thus has more advantages than the absolute conclusion form in terms of problem interpretability. The method is characterized by the generation of evaluation conclusions with probabilistic (reliability) information and thus has more advantages than the absolute conclusion form in terms of problem interpretation. Because of the independence of the stochastic simulation solution method, it is applied to the "bottom-up" evaluation model as an example, and a novel autonomous evaluation method is constructed. Finally, the application of the stochastic simulation evaluation model is illustrated by an example and compared with the absolute form evaluation. The evaluation model is an extension of the traditional evaluation model, which can further broaden the practical application of comprehensive evaluation theory. [ABSTRACT FROM AUTHOR]

The teaching management department carries all the work related to teaching in the whole school. A scientific, efficient, and complete teaching management system cannot only help the teaching management department improve work efficiency and quality but also greatly reduce many problems caused by manual labour risk. This paper designs and implements a teaching management system based on an improved association rule algorithm. First, aiming at the low efficiency of the Apriori algorithm for mining association rules, an association rule model based on interest is proposed. Second, use the MapReduce calculation model to partition the transaction database, then use the improved Apriori optimization algorithm for mining, and finally merge the mining results to obtain frequent itemsets. Through experiments, the optimized algorithm has greatly improved selection mining and computing time than traditional algorithms. [ABSTRACT FROM AUTHOR]

The diversities in the natural, social, and human environment of various regions often lead to the differences in their regional economic development level and industrial structure and layout; moderate regional differences can mobilize economic vitality and improve development efficiency, but excessive differences may lead to social instability or even turbulence. Intelligent algorithms or their improved and hybrid algorithms can recently achieve more suitable solutions to practical problems of nonlinear, discrete, nondifferentiable, and multiple constraints. Therefore, this paper's main point is on the analysis of regional economic development differences based on intelligent hybrid algorithms. On the basis of summarizing and analyzing previous research works, this paper expounded the research status and significance of regional economic development differences analysis, elaborated the development background, current status, and future challenges of intelligent hybrid algorithms, introduced the methods and principles of principal component distance weighting algorithm and sequential quadratic programming algorithm, constructed a basic model for regional economic development differences based on intelligent hybrid algorithm, conducted the construction of analysis index system and the hybrid of intelligence algorithm, analyzed the regional economic development differences based on intelligent hybrid algorithm, performed the principal component analysis and time-space change analysis of regional economic development, and finally carried out a simulation experiment and its result analysis. The study results show that the intelligent hybrid algorithm can simulate the intelligent behavior of certain groups in nature when analyzing the differences in regional economic development, so that it has essentially parallelism, high accuracy, and convergence speed. The study results of this paper provide a reference for further researches on the regional economic development differences analysis based on intelligent hybrid algorithm. [ABSTRACT FROM AUTHOR]

To better reflect the precise clustering results of the data samples with different shapes and densities for affinity propagation clustering algorithm (AP), an improved integrated clustering learning strategy based on three-stage affinity propagation algorithm with density peak optimization theory (DPKT-AP) was proposed in this paper. DPKT-AP combined the ideology of integrated clustering with the AP algorithm, by introducing the density peak theory and k-means algorithm to carry on the three-stage clustering process. In the first stage, the clustering center point was selected by density peak clustering. Because the clustering center was surrounded by the nearest neighbor point with lower local density and had a relatively large distance from other points with higher density, it could help the k-means algorithm in the second stage avoiding the local optimal situation. In the second stage, the k-means algorithm was used to cluster the data samples to form several relatively small spherical subgroups, and each of subgroups had a local density maximum point, which is called the center point of the subgroup. In the third stage, DPKT-AP used the AP algorithm to merge and cluster the spherical subgroups. Experiments on UCI data sets and synthetic data sets showed that DPKT-AP improved the clustering performance and accuracy for the algorithm. [ABSTRACT FROM AUTHOR]

Aiming at the shortcomings of standard particle swarm optimization (PSO) algorithms that easily fall into local optimum, this paper proposes an optimization algorithm (LTQPSO) that improves quantum behavioral particle swarms. Aiming at the problem of premature convergence of the particle swarm algorithm, the evolution speed of individual particles and the population dispersion are used to dynamically adjust the inertia weights to make them adaptive and controllable, thereby avoiding premature convergence. At the same time, the natural selection method is introduced into the traditional position update formula to maintain the diversity of the population, strengthen the global search ability of the LTQPSO algorithm, and accelerate the convergence speed of the algorithm. The improved LTQPSO algorithm is applied to landscape trail path planning, and the research results prove the effectiveness and feasibility of the algorithm. [ABSTRACT FROM AUTHOR]

Federated learning (FL) is a distributed machine learning (ML) framework. In FL, multiple clients collaborate to solve traditional distributed ML problems under the coordination of the central server without sharing their local private data with others. This paper mainly sorts out FLs based on machine learning and deep learning. First of all, this paper introduces the development process, definition, architecture, and classification of FL and explains the concept of FL by comparing it with traditional distributed learning. Then, it describes typical problems of FL that need to be solved. On the basis of classical FL algorithms, several federated machine learning algorithms are briefly introduced, with emphasis on deep learning and classification and comparisons of those algorithms are carried out. Finally, this paper discusses possible future developments of FL based on deep learning. [ABSTRACT FROM AUTHOR]

A terahertz (THz) frequency-modulated continuous wave (FMCW) imaging radar system is developed for high-resolution 3D imaging recently. Aiming at the problems of long data acquisition periods and large sample sizes for the developed imaging system, an algorithm based on compressed sensing is proposed for THz FMCW radar 3D imaging in this paper. Firstly, the FMCW radar signal model is built, and the conventional range migration algorithm is introduced for THz FMCW radar imaging. Then, compressed sensing is extended for THz FMCW radar 3D imaging, and the Newton smooth L0-norm (NSL0) algorithm is presented for sparse measurement data reconstruction. Both simulation and measurement experiments demonstrate the feasibility of reconstructing THz images from measurements even at the sparsity rate of 20%. [ABSTRACT FROM AUTHOR]

As the number of students in universities continues to grow, the university academic management system has a large amount of data on student performance. However, the utilization of these data is only limited to simple query and statistical work, and there is no precedent of using these data for improving English teaching mode. With the application of fuzzy theory in machine learning and artificial intelligence, the fuzzy decision tree algorithm was born by integrating fuzzy set theory with decision tree algorithm. In this paper, we propose a way to obtain the centroids of continuous attribute clustering by K-means algorithm and combine the triangular fuzzy number to fuzzy the continuous data. In addition, this paper analyzes the influence of nearest neighbor distance on classification, introduces Gaussian weight function, gives different voting weights to the neighborhood according to the distance, and establishes a weighted K-nearest neighbor classification algorithm. To address the problem of low classification efficiency of K-nearest neighbor algorithm when the dataset is large, this paper further improves the algorithm and establishes the partitioned weighted K-nearest neighbor algorithm. The classification time was shortened from 11.39 seconds to 5.22 seconds, and the classification efficiency greatly improved. [ABSTRACT FROM AUTHOR]

In order to improve the surveillance effect of the aviation target surveillance radar, this paper improves the traditional filtering algorithm and builds the channel optimization system of the ADS-B aviation target surveillance radar based on the improved filtering algorithm. Moreover, this paper uses algorithm improvement to ensure the positive definite or semipositive definiteness of the state covariance and uses the root mean square volume Kalman filter to avoid the filter divergence or tracking interruption caused by the nonpositive definiteness of the matrix; the filtering principle of the interactive multimodel is to use multiple filters for parallel processing and achieve the adaptive adjustment algorithm residual error by adjusting the one-step prediction covariance in the adjustment algorithm. In addition, this paper combines the actual needs to construct a system functional structure to optimize the channel of the ADS-B aviation target surveillance radar and uses software engineering methods to model and analyze the requirements. Finally, this paper designs experiments to verify system performance. The research results show that the performance of the system constructed in this paper meets actual needs. [ABSTRACT FROM AUTHOR]

The selection of positions for unified power flow controller (UPFC) placement in transmission network is an essential factor, which aids in operating the system in a more reliable and secured manner. This paper focuses on strengthening the power system performance through UPFC placement employing self-adaptive firefly algorithm (SAFA), which selects the best positions along with parameters for UPFC placement. Three single objectives of real power loss reduction, voltage profile improvement, and voltage stability enhancement are considered in this work. IEEE 14, 30, and 57 test systems are selected to accomplish the simulations and to reveal the efficacy of the proposed SAFA approach; besides, solutions are compared with two other algorithms solutions of honey bee algorithm (HBA) and bacterial foraging algorithm (BFA). The proposed SAFA contributes real power loss reduction, voltage profile improvement, and voltage stability enhancement by optimally choosing the placement for UPFC. [ABSTRACT FROM AUTHOR]

To achieve the high-performance control of the surface-mounted permanent magnet synchronous motor (SPMSM) speed control system, this paper proposes a high-order sliding mode control strategy based on a new super twisting algorithm (NSTA). This strategy introduces an adaptive term in its proportional term based on the original super twisting algorithm, which solves low reaching speed and poor antidisturbance ability due to the square root calculation of proportional term in the original super twisting algorithm. The simulation results show that the proposed strategy can effectively improve the system's response speed and antidisturbance and greatly suppress the chattering phenomenon of traditional sliding mode control. [ABSTRACT FROM AUTHOR]

High Utility Itemset Mining (HUIM) is one of the most investigated tasks of data mining. It has broad applications in domains such as product recommendation, market basket analysis, e-learning, text mining, bioinformatics, and web click stream analysis. Insights from such pattern analysis provide numerous benefits, including cost cutting, improved competitive advantage, and increased revenue. However, HUIM methods may discover misleading patterns as they do not evaluate the correlation of extracted patterns. As a consequence, a number of algorithms have been proposed to mine correlated HUIs. These algorithms still suffer from the issue of the computational cost in terms of both time and memory consumption. This paper presents an algorithm, named Efficient Correlated High Utility Pattern Mining (ECoHUPM), to efficiently mine the high utility patterns having strong correlation items. A new data structure based on utility tree (UTtree) named CoUTlist is proposed to store sufficient information for mining the desired patterns. Three pruning properties are introduced to reduce the search space and improve the mining performance. Experiments on sparse, very sparse, dense, and very dense datasets indicate that the proposed ECoHUPM algorithm is efficient as compared to the state-of-the-art CoHUIM and CoHUI-Miner algorithms in terms of both time and memory consumption. [ABSTRACT FROM AUTHOR]

In this paper, we study the automatic construction and extraction of feature variables of sports moments and construct the extraction of the specific variables by artificial intelligence. In this paper, support vector machines, which have better performance in the case of small samples, are selected as classifiers, and multiclass classifiers are constructed in a one-to-one manner to achieve the classification and recognition of human sports postures. The classifier for a single decomposed action is constructed to transform the automatic description problem of free gymnastic movements into a multilabel classification problem. With the increase in the depth of the feature extraction network, the experimental effect is enhanced; however, the two-dimensional convolutional neural network loses temporal information when extracting features, so the three-dimensional convolutional network is used in this paper for spatial-temporal feature extraction of the video. The extracted features are binary classified several times to achieve the goal of multilabel classification. To form a comparison experiment, the results of the classification are randomly combined into a sentence and compared with the results of the automatic description method to verify the effectiveness of the method. The multiclass classifier constructed in this paper is used for human motion pose classification and recognition tests, and the experimental results show that the human motion pose recognition algorithm based on multifeature fusion can effectively improve the recognition accuracy and perform well in practical applications. [ABSTRACT FROM AUTHOR]

This paper addresses a new machine learning-based behavioral strategy using the deep Q-learning algorithm for the RoboCode simulation platform. According to this strategy, a new model is proposed for the RoboCode platform, providing an environment for simulated robots that can be programmed to battle against other robots. Compared to Atari Games, RoboCode has a fairly wide set of actions and situations. Due to the challenges of training a CNN model for such a continuous action space problem, the inputs obtained from the simulation environment were generated dynamically, and the proposed model was trained by using these inputs. The trained model battled against the predefined rival robots of the environment (standard robots) by cumulatively benefiting from the experience of these robots. The comparison between the proposed model and standard robots of RoboCode Platform was statistically verified. Finally, the performance of the proposed model was compared with machine learning based-customized robots (community robots). Experimental results reveal that the proposed model is mostly superior to community robots. Therefore, the deep Q-learning-based model has proven to be successful in such a complex simulation environment. It should also be noted that this new model facilitates simulation performance in adaptive and partially cluttered environments. [ABSTRACT FROM AUTHOR]

This paper addresses the robust constrained model predictive control (MPC) for Takagi-Sugeno (T-S) fuzzy uncertain quantized system with random data loss. To deal with the quantization error and the data loss over the networks, the sector bound approach and the Bernoulli process are introduced, respectively. The fuzzy controller and new conditions for stability, which are written as the form of linear matrix inequality (LMI), are presented based on nonparallel distributed compensation (non-PDC) control law and an extended nonquadratic Lyapunov function, respectively. In addition, slack and collection matrices are provided for reducing the conservativeness. Based on the obtained stability results, a model predictive controller which explicitly considers the input and state constraints is synthesized by minimizing an upper bound of the worst-case infinite horizon quadratic cost function. The developed MPC algorithm can guarantee the recursive feasibility of the optimization problem and the stability of closed-loop system simultaneously. Finally, the simulation example is given to illustrate the effectiveness of the proposed technique. [ABSTRACT FROM AUTHOR]

According to the actual application system model which has bias, this paper analyzes the shortage of the conventional augmented algorithm, the two-stage cubature Kalman filtering algorithm, which is presented on the basis of a two-stage nonlinear transformation. The core ideas of the algorithm are to obtain the block diagonalization of the covariance matrix using the matrix transformation and avoid calculating the covariance of the state and bias to reduce the amount of calculation and ensure a smooth filtering process. Then, the equivalence of the two-stage cubature Kalman filtering algorithm and the cubature Kalman filtering algorithm is proved by updating equivalent transformation. Through the experiment of trajectory tracking of a wheeled robot, it is verified that the two-stage cubature Kalman filtering algorithm can obtain good tracking accuracy and stability with the presence of unknown random bias. Simultaneously, the equivalence of the two-stage cubature Kalman filtering algorithm and cubature Kalman filtering algorithm is verified again using the contrast experiment. [ABSTRACT FROM AUTHOR]

As the world keeps advancing, the need for automated interconnected devices has started to gain significance; to cater to the condition, a new concept Internet of Things (IoT) has been introduced that revolves around smart devicesʼ conception. These smart devices using IoT can communicate with each other through a network to attain particular objectives, i.e., automation and intelligent decision making. IoT has enabled the users to divide their household burden with machines as these complex machines look after the environment variables and control their behavior accordingly. As evident, these machines use sensors to collect vital information, which is then the complexity analyzed at a computational node that then smartly controls these devicesʼ operational behaviors. Deep learning-based guessing attack protection algorithms have been enhancing IoT security; however, it still has a critical challenge for the complex industries' IoT networks. One of the crucial aspects of such systems is the need to have a significant training time for processing a large dataset from the networkʼs previous flow of data. Traditional deep learning approaches include decision trees, logistic regression, and support vector machines. However, it is essential to note that this convenience comes with a price that involves security vulnerabilities as IoT networks are prone to be interfered with by hackers who can access the sensor/communication data and later utilize it for malicious purposes. This paper presents the experimental study of cryptographic algorithms to classify the types of encryption algorithms into the asymmetric and asymmetric encryption algorithm. It presents a deep analysis of AES, DES, 3DES, RSA, and Blowfish based on timing complexity, size, encryption, and decryption performances. It has been assessed in terms of the guessing attack in real-time deep learning complex IoT applications. The assessment has been done using the simulation approach and it has been tested the speed of encryption and decryption of the selected encryption algorithms. For each encryption and decryption, the tests executed the same encryption using the same plaintext for five separate times, and the average time is compared. The key size used for each encryption algorithm is the maximum bytes the cipher can allow. To the comparison, the average time required to compute the algorithm by the three devices is used. For the experimental test, a set of plaintexts is used in the simulation—password-sized text and paragraph-sized text—that achieves target fair results compared to the existing algorithms in real-time deep learning networks for IoT applications. [ABSTRACT FROM AUTHOR]

Constrained optimization plays an important role in many decision-making problems and various real-world applications. In the last two decades, various evolutionary algorithms (EAs) were developed and still are developing under the umbrella of evolutionary computation. In general, EAs are mainly categorized into nature-inspired and swarm-intelligence- (SI-) based paradigms. All these developed algorithms have some merits and also demerits. Particle swarm optimization (PSO), firefly algorithm, ant colony optimization (ACO), and bat algorithm (BA) have gained much popularity and they have successfully tackled various test suites of benchmark functions and real-world problems. These SI-based algorithms follow the social and interactive principles to perform their search process while approximating solution for the given problems. In this paper, a multiswarm-intelligence-based algorithm (MSIA) is developed to cope with bound constrained functions. The suggested algorithm integrates the SI-based algorithms to evolve population and handle exploration versus exploitation issues. Thirty bound constrained benchmark functions are used to evaluate the performance of the proposed algorithm. The test suite of benchmark function is recently designed for the special session of EAs competition in IEEE Congress on Evolutionary Computation (IEEE-CEC′13). The suggested algorithm has approximated promising solutions with good convergence and diversity maintenance for most of the used bound constrained single optimization problems. [ABSTRACT FROM AUTHOR]

Digital image data processing is mainly to input digital image data into a computer to complete the conversion of a continuous spatially distributed image model into a discrete digital model so that the computer can identify, process, and store the processing process of digital image information. Geographic information system (GIS) is a computer system that integrates multiple forms of information expression, and it integrates functions such as collection, processing, transmission, storage, management, analysis, expression, and query retrieval, which can quickly discover the spatial distribution of things and their attributes and can express the results accurately and vividly in various intuitive forms. Therefore, on the basis of summarizing and analyzing previous research works, this paper expounded the research status and significance of processing methods for digital image data, elaborated the development background, current status, and future challenges of the GIS technology, introduced the methods and principles of permutation matrix algorithm and subimage averaging method, constructed the processing model for digital image data based on GIS, analyzed the data structure and its database establishment for digital image, proposed the processing methods for digital image data based on GIS, performed the enhancement processing and calculation classification of digital image data, and finally conducted a case analysis and its result discussion. The study results show that the proposed processing methods for digital image data based on GIS can perform analogue-to-digital conversion of continuous images, complete the steps of sampling, layering, and quantization, and then encode the obtained discrete digital signal into the computer to form an in-plane collection of pixels; this processing method can also organically combine spatial information and image data and identify, process, and store digital image data from both spatial and attribute aspects. The study results of this paper provide a reference for further research on the processing methods for digital image data based on GIS. [ABSTRACT FROM AUTHOR]

The traditional image stitching method has some shortcomings such as double shadow, chromatic aberration, and stitching. In view of this, this paper proposes a power function-weighted image stitching method that combines SURF optimization and improved cell acceleration. First, the method uses the cosine similarity to preliminarily judge the similarity of the feature points and then uses the two-way consistency mutual selection to filter the feature point pairs again. Simultaneously, some incorrect matching points in the reverse matching are eliminated. Finally, the method uses the MSAC algorithm to perform fine matching. Then, the power function-weighted fusion algorithm is used to calculate the weight of the center point. The power function weight of the accelerated cell is used to perform the final image fusion. The experimental results show that the matching accuracy rate of the proposed method is about 11 percentage points higher than the traditional SURF algorithm, and the time is reduced by about 1.6 s. In the image fusion stage, this paper first selects images with different brightness, angles, resolutions, and scales to verify the effectiveness of the proposed method. The results show that the proposed method effectively solves the ghosting and stitching seams. Comparing with the traditional fusion algorithm, the time consumption is reduced by at least 2 s, the mean square error is reduced by about 1.32%∼1.48%, and the information entropy is improved by about 0.98%∼1.70%. The proposed method has better performance in matching accuracy and fusion effect and has better stitching quality. [ABSTRACT FROM AUTHOR]

This paper concentrates on the predefined-time trajectory tracking for an uncertain robotic manipulator system. First, a modified predefined-time control (PTC) algorithm is proposed. Subsequently, with the help of proposed modified PTC algorithm and the nonsingular design method of terminal sliding mode, a novel nonsingular terminal sliding-mode control (NTSMC) scheme is proposed for ensuring the predefined-time convergence of tracking errors. The advantages of the newly proposed control scheme are as follows. (i) Unlike the conventional predefined-time sliding-mode control (SMC) which only guarantees the predefined-time convergence of sliding-mode surface, the proposed scheme can guarantee the predefined-time convergence of tracking errors. (ii) Compared with the conventional PTC algorithm, the proposed modified PTC algorithm can reduce the initial control peaking and enhance the precision of convergence time. The performance and effectiveness of the proposed control scheme are illustrated by comparing with the existing methods. [ABSTRACT FROM AUTHOR]

Resource saving has become an integral aspect of manufacturing in industry 4.0. This paper proposes a multisystem optimization (MSO) algorithm, inspired by implicit parallelism of heuristic methods, to solve an integrated production scheduling with resource saving problem in textile printing and dyeing. First, a real-world integrated production scheduling with resource saving is formulated as a multisystem optimization problem. Then, the MSO algorithm is proposed to solve multisystem optimization problems that consist of several coupled subsystems, and each of the subsystems may contain multiple objectives and multiple constraints. The proposed MSO algorithm is composed of within-subsystem evolution and cross-subsystem migration operators, and the former is to optimize each subsystem by excellent evolution operators and the later is to complete information sharing between multiple subsystems, to accelerate the global optimization of the whole system. Performance is tested on a set of multisystem benchmark functions and compared with improved NSGA-II and multiobjective multifactorial evolutionary algorithm (MO-MFEA). Simulation results show that the MSO algorithm is better than compared algorithms for the benchmark functions studied in this paper. Finally, the MSO algorithm is successfully applied to the proposed integrated production scheduling with resource saving problem, and the results show that MSO is a promising algorithm for the studied problem. [ABSTRACT FROM AUTHOR]

On the basis of ecological footprint theory and tourism ecological footprint theory, the sustainable development indexes such as ecological footprint, ecological carrying capacity, ecological deficit, and ecological surplus of the research area were calculated and the long-term change pattern of each index was analyzed. This paper shows that the ecological footprint of the research area increases year by year, but the ecological footprint is always smaller than the ecological carrying capacity, indicating that the area is still in the state of sustainable development. However, the per capita ecological surplus shows a decreasing trend year by year, indicating that the sustainable development of the region is getting worse. This paper proposes a reordering method of tourist attractions based on heterogeneous information fusion, and realizes the retrieval and reordering of tourist attractions based on user query and fusion of heterogeneous information, so as to help users make travel decisions. In view of the shortage of tourism commercial websites to passively provide scenic spot information, this paper puts forward a scenic spot retrieval method based on query words to enable users to obtain scenic spot information according to their needs, and constructs a tourist consumer data analysis system. The preprocessing methods and methods adopted by the data preprocessing module are analyzed in detail, and the algorithms used in the travel route analysis and consumer spending ability analysis are described in detail. The data of tourism consumers are analyzed by this system, and the results are evaluated. [ABSTRACT FROM AUTHOR]

Let G V , E be a simple undirected complete graph with vertex and edge sets V and E , respectively. In this paper, we consider the degree-constrained k -minimum spanning tree (DC k MST) problem which consists of finding a minimum cost subtree of G formed with at least k vertices of V where the degree of each vertex is less than or equal to an integer value d ≤ k − 2. In particular, in this paper, we consider degree values of d ∈ 2,3 . Notice that DC k MST generalizes both the classical degree-constrained and k -minimum spanning tree problems simultaneously. In particular, when d = 2 , it reduces to a k -Hamiltonian path problem. Application domains where DC k MST can be adapted or directly utilized include backbone network structures in telecommunications, facility location, and transportation networks, to name a few. It is easy to see from the literature that the DC k MST problem has not been studied in depth so far. Thus, our main contributions in this paper can be highlighted as follows. We propose three mixed-integer linear programming (MILP) models for the DC k MST problem and derive for each one an equivalent counterpart by using the handshaking lemma. Then, we further propose ant colony optimization (ACO) and variable neighborhood search (VNS) algorithms. Each proposed ACO and VNS method is also compared with another variant of it which is obtained while embedding a Q-learning strategy. We also propose a pure Q-learning algorithm that is competitive with the ACO ones. Finally, we conduct substantial numerical experiments using benchmark input graph instances from TSPLIB and randomly generated ones with uniform and Euclidean distance costs with up to 400 nodes. Our numerical results indicate that the proposed models and algorithms allow obtaining optimal and near-optimal solutions, respectively. Moreover, we report better solutions than CPLEX for the large-size instances. Ultimately, the empirical evidence shows that the proposed Q-learning strategies can bring considerable improvements. [ABSTRACT FROM AUTHOR]

Road link speed is one of the important indicators for traffic states. In order to incorporate the spatiotemporal dynamics and correlation characteristics of road links into speed prediction, this paper proposes a method based on LDA and GCN. First, we construct a trajectory dataset from map-matched GPS location data of taxis. Then, we use the LDA algorithm to extract the semantic function vectors of urban zones and quantify the spatial dynamic characteristics of road links based on taxi trajectories. Finally, we add semantic function vectors to the dataset and train a graph convolutional network to learn the spatial and temporal dependencies of road links. The learned model is used to predict the future speed of road links. The proposed method is compared with six baseline models on the same dataset generated by GPS equipped on taxis in Shenzhen, China, and the results show that our method has better prediction performance when semantic zoning information is added. Both composite and single-valued semantic zoning information can improve the performance of graph convolutional networks by 6.46% and 8.35%, respectively, while the baseline machine learning models work only for single-valued semantic zoning information on the experimental dataset. [ABSTRACT FROM AUTHOR]

The power converters are widely used in several industrial applications where it is necessary to obtain from a fixed voltage another one higher or lower than the original. In this paper, we focus on the DC-DC (direct current) boost converters, where to guarantee the desired voltage, an internal current tracking loop is usually used. However, this tracking cannot be assured in the presence of unknown load changes and external perturbations when traditional controller strategies are implemented. In this paper, an advanced control strategy is proposed to ensure the current tracking using a saturated super-twisting controller on the power converter. The finite-time current tracking of a DC-DC boost converter is assured in the presence of bounded Lipschitz perturbations composed by unknown load changes and exogenous signals. The proposed approach generates a continuous bounded control signal applied to the converter by using a sigma-delta modulator Σ Δ M . The controller gains are tuned to obtain finite-time stabilization of the tracking error, while the control signal remains bounded. To illustrate the effectiveness of the proposed results, the controller is applied to a physical boost converter using the hardware implemented Σ Δ M and an STM32 Discovery development card. Besides, the controller is compared with a first-order sliding mode controller showing that for small sample times, the energy of the error signal is reduced. [ABSTRACT FROM AUTHOR]

Mobile manipulators are widely used in different fields for transferring and grasping tasks such as in medical assisting devices, industrial production, and hotel services. It is challenging to improve navigation accuracies and grasping success rates in complex environments. In this paper, we develop a multisensor-based mobile grasping system which is configured with a vision system and a novel gripper set in an UR5 manipulator. Additionally, an error term of a cost function based on DWA (dynamic window approach) is proposed to improve the navigation performance of the mobile platform through visual guidance. In the process of mobile grasping, the size and position of the object can be identified by a visual recognition algorithm, and then the finger space and chassis position can be automatically adjusted; thus, the object can be grasped by the UR5 manipulator and gripper. To demonstrate the proposed methods, comparison experiments are also conducted using our developed mobile grasping system. According to the analysis of the experimental results, the motion accuracy of the mobile chassis has been improved significantly, satisfying the requirements of navigation and grasping success rates, as well as achieving a high performance over a wide grasping size range from 1.7 mm to 200 mm. [ABSTRACT FROM AUTHOR]

Density peaks clustering (DPC) is an advanced clustering technique due to its multiple advantages of efficiently determining cluster centers, fewer arguments, no iterations, no border noise, etc. However, it does suffer from the following defects: (1) difficult to determine a suitable value of its crucial cutoff distance parameter, (2) the local density metric is too simple to find out the proper center(s) of the sparse cluster(s), and (3) it is not robust that parts of prominent density peaks are remotely assigned. This paper proposes improved density peaks clustering based on natural neighbor expanded group (DPC-NNEG). The cores of the proposed algorithm contain two parts: (1) define natural neighbor expanded (NNE) and natural neighbor expanded group (NNEG) and (2) divide all NNEGs into a goal number of sets as the final clustering result, according to the closeness degree of NNEGs. At the same time, the paper provides the measurement of the closeness degree. We compared the state of the art with our proposal in public datasets, including several complex and real datasets. Experiments show the effectiveness and robustness of the proposed algorithm. [ABSTRACT FROM AUTHOR]

This paper is in view of the current algorithm to solve Multiple-Devices-Process integrated scheduling problems without considering parallel process between the parallel processing. Meanwhile, the influence of the first processing procedure on the second processing procedure is ignored, resulting in poor tightness between serial processes and poor parallelism between parallel processes, ultimately affecting the product scheduling results. We proposed Multiple-Devices-Process integrated scheduling algorithm with time-selective for process sequence. The proposed Multiple-Devices-Process sequencing strategy determines the process of scheduling order and improves the tightness between serial process. This paper presents a method to determine the quasischeduling time point of the multiequipment process, the time-selective strategy of Multiple-Devices-Process, and the time-selective adjustment strategy of Multiple-Devices-Process so that the first and the second processing processes cooperate with each other, and the purpose of improving the tightness of the serial process and the parallelism of the parallel processes is achieved, so as to shorten the product processing time. [ABSTRACT FROM AUTHOR]

In this paper, a novel multiobjective lightning flash algorithm (MOLFA) is proposed to solve the multiobjective optimization problem. The charge population state of the lightning flash algorithm is defined, and we prove that the charge population state sequence is a Markov chain. Since the convergence analysis of MOLFA is to investigate whether a Pareto optimal solution can be reached when the optimal charge population state is obtained, the development of a charge population state is analyzed to achieve the goal of this paper. Based on the martingale theory, the MOLFA convergence analysis is carried out in terms of the supermartingale convergence theorem, which shows that the MOLFA can reach the global optimum with probability one. Finally, the effectiveness of the proposed MOLFA is verified by a numerical simulation example. [ABSTRACT FROM AUTHOR]

Human gait phase detection is a significance technology for robotics exoskeletons control and exercise rehabilitation therapy. Inertial Measurement Units (IMUs) with accelerometer and gyroscope are convenient and inexpensive to collect gait data, which are often used to analyze gait dynamics for personal daily applications. However, current deep-learning methods that extract spatial and the isolated temporal features can easily ignore the correlation that may exist in the high-dimensional space, which limits the recognition effect of a single model. In this study, an effective hybrid deep-learning framework based on Gaussian probability fusion of multiple spatiotemporal networks (GFM-Net) is proposed to detect different gait phases from multisource IMU signals. Furthermore, it first employs the gait information acquisition system to collect IMU data fixed on lower limb. With the data preprocessing, the framework constructs a spatial feature extractor with AutoEncoder and CNN modules and a multistream temporal feature extractor with three collateral modules combining RNN, LSTM, and GRU modules. Finally, the novel Gaussian probability fusion module optimized by the Expectation-Maximum (EM) algorithm is developed to integrate the different feature maps output by the three submodels and continues to realize gait recognition. The framework proposed in this paper implements the inner loop that also contains the EM algorithm in the outer loop and optimizes the reverse gradient in the entire network. Experiments show that this method has better performance in gait classification with accuracy reaching more than 96.7%. [ABSTRACT FROM AUTHOR]

Software refactoring is a software maintenance action to improve the software internal quality without changing its external behavior. During the maintenance process, structural refactoring is performed by remodularizing the source code. Software clustering is a modularization technique to remodularize artifacts of source code aiming to improve readability and reusability. Due to the NP hardness of the clustering problem, evolutionary approaches such as the genetic algorithm have been used to solve this problem. In the structural refactoring literature, there exists no search-based algorithm that employs a hierarchical approach for modularization. Utilizing global and local search strategies, in this paper, a new search-based top-down hierarchical clustering approach, named TDHC, is proposed that can be used to modularize the system. The output of the algorithm is a tree in which each node is an artifact composed of all artifacts in its subtrees and is a candidate to be a software module (i.e., cluster). This tree helps a software maintainer to have better vision on source code structure to decide appropriate composition points of artifacts aiming to create modules (i.e., files, packages, and components). Experimental results on seven folders of Mozilla Firefox with different functionalities and five other software systems show that the TDHC produces modularization closer to the human expert's decomposition (i.e., directory structure) than the other existing algorithms. The proposed algorithm is expected to help a software maintainer for better remodularization of a source code. The source codes and dataset related to this paper can be accessed at https://github.com/SoftwareMaintenanceLab. [ABSTRACT FROM AUTHOR]

This paper proposes a new manifold-based dimension reduction algorithm framework. It can deal with the dimension reduction problem of data with noise and give the dimension reduction results with the deviation values caused by noise interference. Commonly used manifold learning methods are sensitive to noise in the data. Mean computation, a denoising method, is an important step in data preprocessing but leads to a loss of local structural information. In addition, it is difficult to measure the accuracy of the dimension reduction of noisy data. Thus, manifold learning methods often transform the data into an approximately smooth manifold structure; however, practical data from the physical world may not meet the requirements. The proposed framework follows the idea of the localization of manifolds and uses graph sampling to determine some local anchor points from the given data. Subsequently, the specific range of localities is determined using graph spectral analysis, and the density within each local range is estimated to obtain the distribution parameters. Then, manifold-based dimension reduction with distribution parameters is established, and the deviation values in each local range are measured and further extended to all data. Thus, our proposed framework gives a measurement method for deviation caused by noise. [ABSTRACT FROM AUTHOR]

Multiobject Tracking (MOT) is one of the most important abilities of autonomous driving systems. However, most of the existing MOT methods only use a single sensor, such as a camera, which has the problem of insufficient reliability. In this paper, we propose a novel Multiobject Tracking method by fusing deep appearance features and motion information of objects. In this method, the locations of objects are first determined based on a 2D object detector and a 3D object detector. We use the Nonmaximum Suppression (NMS) algorithm to combine the detection results of the two detectors to ensure the detection accuracy in complex scenes. After that, we use Convolutional Neural Network (CNN) to learn the deep appearance features of objects and employ Kalman Filter to obtain the motion information of objects. Finally, the MOT task is achieved by associating the motion information and deep appearance features. A successful match indicates that the object was tracked successfully. A set of experiments on the KITTI Tracking Benchmark shows that the proposed MOT method can effectively perform the MOT task. The Multiobject Tracking Accuracy (MOTA) is up to 76.40% and the Multiobject Tracking Precision (MOTP) is up to 83.50%. [ABSTRACT FROM AUTHOR]

With the rapid development of market economy, the task scheduling model has become the core problem in the field of corporate management. In order to solve the problem that the stability of the model will decline due to the interference of human factors in the process of multitask scheduling in the traditional algorithm, a reasonable scheduling model based on the priority principle is proposed in this paper. This paper expounds the principle of multitask scheduling in company management, constructs a multitask scheduling network model based on priority, performs virtual scheduling on the model, searches the optimal solution in the solution space of the scheduling problem, and obtains the reasonable multitask scheduling method in the company management. Through the analysis of relevant simulation experiments, it can be concluded that the task scheduling algorithm proposed in this paper can not only allocate the corresponding resources for high-priority tasks effectively, but also can save the cost, so as to show better execution efficiency. [ABSTRACT FROM AUTHOR]

The grasshopper optimization algorithm (GOA) is a metaheuristic algorithm that mathematically models and simulates the behavior of the grasshopper swarm. Based on its flexible, adaptive search system, the innovative algorithm has an excellent potential to resolve optimization problems. This paper introduces an enhanced GOA, which overcomes the deficiencies in convergence speed and precision of the initial GOA. The improved algorithm is named MOLGOA, which combines various optimization strategies. Firstly, a probabilistic mutation mechanism is introduced into the basic GOA, which makes full use of the strong searchability of Cauchy mutation and the diversity of genetic mutation. Then, the effective factors of grasshopper swarm are strengthened by an orthogonal learning mechanism to improve the convergence speed of the algorithm. Moreover, the application of probability in this paper greatly balances the advantages of each strategy and improves the comprehensive ability of the original GOA. Note that several representative benchmark functions are used to evaluate and validate the proposed MOLGOA. Experimental results demonstrate the superiority of MOLGOA over other well-known methods both on the unconstrained problems and constrained engineering design problems. [ABSTRACT FROM AUTHOR]

Trams have increasingly deployed object detectors to perceive running conditions, and deep learning networks have been widely adopted by those detectors. Growing neural networks have incurred severe attacks such as adversarial example attacks, imposing threats to tram safety. Only if adversarial attacks are studied thoroughly, researchers can come up with better defence methods against them. However, most existing methods of generating adversarial examples have been devoted to classification, and none of them target tram environment perception systems. In this paper, we propose an improved projected gradient descent (PGD) algorithm and an improved Carlini and Wagner (C&W) algorithm to generate adversarial examples against Faster R-CNN object detectors. Experiments verify that both algorithms can successfully conduct nontargeted and targeted white-box digital attacks when trams are running. We also compare the performance of the two methods, including attack effects, similarity to clean images, and the generating time. The results show that both algorithms can generate adversarial examples within 220 seconds, a much shorter time, without decrease of the success rate. [ABSTRACT FROM AUTHOR]

High-dimensional and unbalanced data anomaly detection is common. Effective anomaly detection is essential for problem or disaster early warning and maintaining system reliability. A significant research issue related to the data analysis of the sensor is the detection of anomalies. The anomaly detection is essentially an unbalanced sequence binary classification. The data of this type contains characteristics of large scale, high complex computation, unbalanced data distribution, and sequence relationship among data. This paper uses long short-term memory networks (LSTMs) combined with historical sequence data; also, it integrates the synthetic minority oversampling technique (SMOTE) algorithm and K-nearest neighbors (kNN), and it designs and constructs an anomaly detection network model based on kNN-SMOTE-LSTM in accordance with the data characteristic of being unbalanced. This model can continuously filter out and securely generate samples to improve the performance of the model through kNN discriminant classifier and avoid the blindness and limitations of the SMOTE algorithm in generating new samples. The experiments demonstrated that the structured kNN-SMOTE-LSTM model can significantly improve the performance of the unbalanced sequence binary classification. [ABSTRACT FROM AUTHOR]

With the development of deep learning, many approaches based on neural networks are proposed for code clone. In this paper, we propose a novel source code detection model At-biLSTM based on a bidirectional LSTM network with a self-attention layer. At-biLSTM is composed of a representation model and a discriminative model. The representation model firstly transforms the source code into an abstract syntactic tree and splits it into a sequence of statement trees; then, it encodes each of the statement trees with a deep-first traversal algorithm. Finally, the representation model encodes the sequence of statement vectors via a bidirectional LSTM network, which is a classical deep learning framework, with a self-attention layer and outputs a vector representing the given source code. The discriminative model identifies the code clone depending on the vectors generated by the presentation model. Our proposed model retains both the syntactics and semantics of the source code in the process of encoding, and the self-attention algorithm makes the classifier concentrate on the effect of key statements and improves the classification performance. The contrast experiments on the benchmarks OJClone and BigCloneBench indicate that At-LSTM is effective and outperforms the state-of-art approaches in source code clone detection. [ABSTRACT FROM AUTHOR]